J. Am. Chem. SOC. 1981,103, 230-232
230
are seen to be somewhat long and the eight cyclopentane rings come in a variety of conformations. Of major interest are the distances between atoms where bonds are needed to complete the dodecahedrane framework. These are C(2)-C(6), 3.27; C( 1)C(20), 3.15; C(7)-C(20), 3.31 A; all are well within achievable range.15 Because it was already obvious that the carbomethoxy group in l l a could not be enticed into chemical reaction by photochemical means, the decision was made to proceed with aldehyde 12. This substance was arrived at in excellent yield by sequential diisobutylaluminum hydride reduction and pyridinium chlorochromate oxidation.16 The fact that the CHO group was bonded to a fully substituted carbon now had to be contended with. The literature dealing with the photochemical cyclization of aldehydes to cycl~butanols'~ leaves no doubt that these structural features are most conducive to decarbonylation. While 12 was certainly prone to carbon monoxide extrusion, a 29% yield of the epimeric "homo-Norrish" cyclopentanols could be realized in deoxygenated toluene-ethanol (9:l) solution at -78 OC (450-W Hanovia lamp, Pyrex). With subsequent pyridinium chlorochromate oxidation,
Ck3
12
Cli3
ck 3
'2
'4
assumed to be caused by the sterically compressed opposed methylene hydrogens. Molecular models of this decaquinane reveal that the opposed methylene groups must experience substantial steric repulsion adequate to force a skewed orientationm Preliminary efforts to uncover a dehydrogenation catalyst effective on 2 or an isomerization catalyst which would cyclize 15 have yet to be successful. While investigations along the preceding lines continue, other workable solutions to this final maneuver and to the removal (or noninstallment) of the vestigial methyl groups are currently being examined.I9 Supplementary Material Available: Final atomic (Table I) and anisotropic thermal parameters (Table 11), bond lengths (Table 111), bond angles (Table IV), and torsion angles (Table V) for l l b (6 pages). Ordering information is given on any current masthead page. t Presidential Fellow, 1979-1 980. (19) The National Institutes of Health (Grant No. AI-11490) provided the financial support which made this research possible. (20) Note Added in Proof: Dr.Peter Engel (Bern) has recently confirmed the structural assignment to 2 by X-ray analysis.
13C N M R Spectra of Carbonium Ions in the Solid State: The sec-Butyl Cation Philip C. Myhre*t and Costantino S . Yannoni* IBM Research Laboratory San Jose, California 95193 Received May 27, 1980
CH; 2
Ck,
15
the diseco ketone 13 was obtained. Furthermore, not only could 13 be photocyclized reproducibly in high yield, but removal of the tertiary hydroxyl group in 14 and saturation of the double bond in 15 were encouragingly simple and efficient steps. With arrival at the beautifully crystalline secododecahedrane 2 (mp 235-240 OC, sublime), a return to C, symmetry materializes, a phenomenon reflected in the appearance of only eight lines in its I3C N M R spectrum.'* The infrared spectrum is characterized by an unusually high absorption at 3 150 cm-I, (14) The crystals were triclinic, space group PI, with u = 8.517 (3), b = 8.738 (2), c = 13.191 (5) A; a = 70.86 (3)', /3 = 80.94 (3)', y = 85.01 (2)'; da,d = 1.279 g cm-3 for Z = 2 (C24H3202, M,352.52). The intensity data were measured with a Hilger-Watts diffractometer (Ni-filtered Cu Ka radiation, 8-20 scans, pulse-height discrimination). A crystal measuring approximately 0.15 X 0.35 X 0.6 mm was used for data collection. A total of 2472 reflections were measured for 8 < 57", of which 2244 were considered to be observed [I> 2Su(I)]. The structure was solved by a multiple solution procedure [Germain, G.; Main, P.; Woolfson, M. M. Acra Crysfullogr.,Secr. A 1971, A27, 3681 and was refined by full-matrix least squares. In the final refinement anisotropic thermal parameters were used for the heavier atoms and isotropic temperature factors were used for the hydrogen atoms. The hydrogen atoms were refined isotropically. The final discrepancy indices are R = 0.045 and w R = 0.059 for the 2244 observed reflections. The final difference map has no peaks greater than k0.2 e A-3. (15) This is particularly so for ketone photochemistry in the solid state where it has been shown that the >=O---Ha-Cf distance should be 62.5 A, the > ( H O ) e - - e t distance 5 3 . 5 A, and the --H,/O=Cf angle approximately 90' with H, 510' out of the carbonyl plane: Scheffer, J. R.; Dzakpasu, A. A. J. Am. Chem. SOC.1978, 100, 2163. (16) Mp 132-134 'C; 'H NMR (6, CDC13) 9.98 (s, 1 H), 3.8-0.7 (series of m, 21 H), 1.20 (s, 3 H), and 1.12 (s, 3 H); m / e calcd: 308.2140; found: 308.2150. (17) (a) Orban, I.; Schaffner, K.;Jeger, 0. J . Am. Chem. SOC.1963,85, 3033. (b) Schaffner, K.Chimiu 1965, 19, 575; (c) Agosta, W. C.; Herron, D. K.J. Am. Chem. SOC.1968,90, 7025. (d) Coyle, J. D. J . Chem. SOC.B 1971,2254. (d) Lischewski, M.; Adam, G.; Serebryakov, E. P. Tetrahedron r m... 1980. 45. _. (18) '3C NMR (ppm, CDCl3) 78.40 (s), 70.15 (d), 68.16 (d), 66.08 (d), 58.99 (d), 52.29 (d), 33.64 (9). and 32.58 (t).
The use of magic angle spinning in combination with N M R pulse techniques allows moderately high resolution I3C N M R spectroscopy in the solid state.' With the inclusion of variable temperature capability: solid-state 13CN M R is a powerful tool for the study of very reactive intermediates. We report procedures that allow preparation of solid samples of thermally unstable carbonium ions and subsequent manipulation of samples for solid-state NMR studies at temperatures close to 77 K.j A study of the sec-butyl cation exemplifies these procedures. 2-Butan01-3-'~C(55 atom % enriched) was prepared from acetic acid-l-13C by a standard sequence, and the alcohol was converted to 2-~hlorobutane.~*~ An intimate mixture of the alkyl chloride (1.5 mmol) and SbF5 (- 13 mmol) was deposited at torr on a cooled surface with the use of a vapor deposition apparatus (Figure 1). Our apparatus and deposition procedures are similar to those described by Saundem6 The O-ring joint on the reactor permits rapid detachment of the reactor base. Concentric jets
-
~~
*Department of Chemistry, Harvey Mudd College, Claremont, CA 91711. (1) Schaefer, J.; Stejskal, E. 0.;Buchdahl, R. Mucromolecules 1977, 10, 384. (2) Fyfe, C. A.; Mossbrugger, H.; Yannoni, C. S.J . Mugn. Reson. 1979, 36, 61. (3) Well-resolved I3C NMR spectra have been reported. for stable carbonium ion salts. (Lyerla, J. R.; Yannoni, C. S.; Bruck, D.; Fyfe, C. A. J . Am. Chem. SOC.1979, 101, 4770.) (4) The sequence used was
1 .
Mn
CH3CH(OH)*CH2CH3
CHXHO
HOCHEHOH
HCI
CH3CHCI*CH2CH3 CH3*CHCICH2CH, (75%) (25%) where * denotes the I3C label. (5) The small volume of the rotors (-60 pL) and the use of excess SbFl necessitated the use of 13C-enrichedsamples to minimize data acquisition time. (6) Saunders, M.; Cox, D.; Lloyd, J. R. J . Am. Chem. SOC.1979, 101, 6656.
0002-7863/81/1503-230$01 .OO/O 0 1980 American Chemical Society
J. Am. Chem. SOC., Vol. 103, No. I, 1981 231
Communications to the Editor Codepotit
Load
/I
-127°C
Figure 1. Sketch of apparatus used for d e p o s i t i o n and loading of solid mixtures of alkyl halide and antimony pentafluoride.
facilitate mixing of reactants. The Dewar jacket improves visual observation, After deposition, the reactor base was detached with a flow of cold nitrogen gas, and a cap equipped with a set of stainless steel tools was attached. A cold nitrogen purge was maintained during sample loading. The tools (Figure 1) consisted of a quick release holder for the rotor, a scoop, a tamping rod, and a combination rotor lid holder and screwdriver. By manipulating these tools, the solid could be transferred to the rotor and the threaded rotor lid secured at temperatures less than -160 '(2.7 The cold rotor is quickly dropped into a cup of liquid nitrogen. Transfer to the stator and spinning are initiated under liquid nitrogen before insertion into the precooled probes2 N M R spectra of the 2-chlorobutane-antimony pentafluoride deposition are shown in Figure 2.' The spectra were obtained with 'H-I3C cross polarization and decoupling of both 19F and 'H. Temperatures reported are those of the exiting propellant gas. The first spectrum, recorded at -85 OC, shows lines characteristic of the sec-butyl cation, represented in conventional form in eq 1, with C-1 and C-4 resonances at 22 ppm and C-2 and C-3 CH3-CH"-CH2-CH3
+ CH3-CH2-CH"-CH3
(1)
resonances a t 170 ppm (Me4,%). These values are close to the solution-state chemical shift^.^ The resonance a t 48 ppm is assigned to the methyl carbons of the tert-butyl cation. At lower temperatures, the same signals were observed with general line broadening and an increase in intensity of the tert-butyl cation signals.I0 Upon warming back to -85 OC, a spectrum identical with the first spectrum in the series was obtained. There was no significant spectral change upon warming to -60 OC. The following conclusions may be drawn from the spectra: (1) Carbonium ions can be formed a t low temperatures in the solid state by interaction of alkyl halides and SbF5." (2) At temperatures of ion formation (